Ask a school why a student left and you will usually get an answer about price or about a competitor.
Then look at the attendance record. Most of the time the leaving started about seven weeks before anyone noticed, and it did not look like leaving. It looked like a busy family.
What the pattern actually looks like
Withdrawal has a shape, and it is remarkably consistent:
- One absence, explained. Illness, a school test, a family trip.
- Two weeks of full attendance. This is the part that fools everyone.
- Two absences close together, one of them unexplained.
- A pattern by day — the Tuesday class specifically, not the Thursday one.
- Silence.
Between step three and step four there is usually a fortnight. That fortnight is the entire opportunity. Before it, there is nothing to act on. After it, the family has already mentally left and a phone call becomes an awkward request to reconsider.
Why nobody catches it
Not carelessness. It is that the information is distributed. The Tuesday teacher notices the gap. The Thursday teacher sees the same child present and cheerful. The front desk sees a parent who is friendly and pays on time. No single person sees the pattern, and each of them individually is right that nothing is wrong.
This is why "our teachers know their students" is not a retention strategy. It is true and insufficient. The signal only exists when you put the two classes side by side.
Two absences is the threshold worth using
Not because two is magic, but because it is the last point at which a call is still cheap and unremarkable.
At two absences you can ask a genuine question — is everything all right, is the level right, is the time still working — and it reads as attention rather than pursuit. At five, the same call reads as a business trying to keep a customer, and the parent hears it that way.
Set the threshold, apply it to everybody, and let the system tell someone. Do not rely on noticing.
What to say
The instinct is to lead with the offer: a make-up class, a discount, a level change. Resist it. Leading with a remedy tells the parent you have already decided what is wrong, and if you guessed incorrectly you have just confirmed that you do not know their child.
Ask first. Most of the answers are boring, and boring is good news — a schedule clash with a school activity, a grandparent who used to do the pickup, a term of exams. All are solvable if you know about them in week three.
The answers that are not boring — the child does not want to come, the class moved past them, something happened with another student — are the ones you would never have found out from a discount offer.
The one that is nobody's fault
Sometimes the answer is that the family is moving, or money got tight, or the child genuinely wants to stop. You cannot save those, and trying hard to is how a school gets a reputation for being difficult to leave.
Retention work is not about persuading people to stay. It is about making sure the ones who left for a fixable reason had somebody notice in time to fix it.
Schoolory flags students who miss several sessions in a row and, on a student record, will rank the risk of dropping out with the reasoning behind it — see the attendance guide for how the flags work.